The 2001 authorization of the No Child Left Behind Act and its standards and accountability requirements generated interest among state education agencies in Florida, Mississippi, and North Carolina, which are served by the Regional Educational Laboratory Southeast, in monitoring changes in student reading and math proficiency at the school level. This study was requested by governing board members representing North Carolina, members of the Improving Literacy Research Alliance (which includes representatives from Florida) and, members of the Improving Schools in Mississippi Research Alliance. All three of these states monitor and report changes in performance for schools. However, this study goes beyond reporting averages to examine school-level academic performance using a growth curve modeling approach. This approach can provide stakeholders with a deeper understanding of trends in student proficiency at the school level, by grade at the school level, and among key demographic groups to allow for more accurate policy responses. This study uses growth curve modeling to investigate trends in student reading and math proficiency on state accountability assessments for grades 3-8. Growth curve modeling makes it possible to determine if growth rates in reading and math proficiency are statistically significant and if the differences in proficiency growth rates between grades and subgroups differ in statistically significant ways. Using four to six years of publicly available school-level data between school years 2007/08 and 2013/14 from each state department of education, this study assessed trends in three areas. First, it estimated trends in average school-level student growth rates in reading and math proficiency on the statewide assessment and examined whether these growth rates varied across grades 3 through 8. Second, it calculated average school-level student growth rates in reading and math proficiency for racial/ethnic subgroups and economic subgroups (eligibility for the federal school lunch program, a proxy for economic disadvantage) in grades 3-8. Third, it examined whether there were any statistically significant decreases in achievement gaps by grade between White and Black students, between White and Hispanic students, and between economic subgroups. This information was then used to estimate reading and math proficiency gaps that remained at the end of the period studied. In general, average school-level student academic proficiency rates increased for most subgroups across grades and subjects in all three states. In addition, reading and math achievement gaps decreased for most subgroups; however, achievement gaps remained large despite the decreases. More findings include: (1) The school-level percentage of students who scored proficient in reading and math increased in all three states--Florida, Mississippi, and North Carolina--over the period studied; (2) School-level growth rates in student reading and math proficiency differed by grade level in each state, though most grades showed a statistically significant increase over the period examined; (3) School-level growth rates in student reading and math proficiency differed across racial/ethnic subgroups, though most subgroups showed a statistically significant increase; (4) The reading and math proficiency achievement gaps between White and Black students and White and Hispanic students decreased in most grades in all three states; however, the gaps still exceeded 10 percentage points; (5) Average school-level proficiency rates in reading and math differed by grade between economic subgroups in Florida and Mississippi, with most grade-economic subgroup combinations showing a significant increase. These data were not available for North Carolina; and (6) The achievement gap in reading and math proficiency between economic subgroups decreased in all grades in Florida and Mississippi; however, the gaps still exceed 13 percentage points. Details on the analyses are appended.
